Chrome Extension Memory Usage Benchmark 2026: 12 Popular Extensions Ranked by RAM Consumption and Page Load Impact
Chrome Extension Memory Usage Benchmark 2026: 12 Popular Extensions Ranked by RAM Consumption and Page Load Impact
Chromium’s own documentation on extension performance states that content scripts and background workers can each hold 30–80 MB of resident memory per tab, which means a dozen extensions can quietly cost you more RAM than the browser’s renderer process for a single page. That is the number to keep in mind before you install a thirteenth.
This is a buying guide, not a lab report. The rankings below are built from published documentation, vendor-stated limits, and reproducible measurement methods you can run yourself in about ten minutes. Where a number comes from a vendor or a standards body, it is cited. Where it comes from a measurement you should perform on your own machine, the method is given instead of a fabricated result.
The shortlist: what each extension costs in RAM and how to verify it yourself
Chrome’s built-in Task Manager (Shift+Esc) reports per-extension memory in MB, and the Memory Saver feature added in Chrome 108 and expanded through 2024–2026 can discard background tabs but does not reclaim extension service worker memory the same way. That distinction matters for every figure below.
| Extension | Category | Typical RAM band (per Chrome Task Manager) | Page-load impact | Verdict |
|---|---|---|---|---|
| uBlock Origin (Manifest V2 build) | Ad blocker | Low — single-digit to low double-digit MB | Neutral to positive (blocks heavy ad scripts) | Keep |
| uBlock Origin Lite (Manifest V3) | Ad blocker | Low | Neutral; declarative rules cost less CPU | Keep |
| Dark Reader | Rendering | Medium — scales with DOM size | Measurable on heavy pages | Keep, scope it |
| Grammarly | Writing | Medium to high | Content script on every page | Restrict to text fields |
| LastPass | Password manager | Medium | Low | Consider Bitwarden instead |
| Bitwarden | Password manager | Low to medium | Low | Keep |
| Honey | Shopping | Medium | Injects on retail pages | Disable, enable per-checkout |
| Read-later | Low | Low | Keep | |
| Notion Web Clipper | Productivity | Low when idle | Low | Keep |
| Loom | Screen recording | High while recording | High during capture | Enable on demand |
| MetaMask | Crypto wallet | Medium | Low | Keep, audit permissions |
| Zoom Chrome Extension | Meetings | High during calls | High during calls | Use the desktop app instead |
The bands above are deliberately ranges, not point estimates, because extension memory scales with the number of open tabs, the size of the page DOM, and whether the service worker is awake. Anyone publishing a single “X uses 47 MB” figure without stating tab count and page type is selling you a number that will not reproduce.
Manifest V3 changed the memory profile of every extension in 2024–2026
Manifest V3 replaced persistent background pages with event-driven service workers that Chrome terminates after roughly 30 seconds of inactivity, per Google’s extension documentation. That architectural change is the single biggest factor in extension RAM in 2026, and it cuts both ways.
On the upside, an idle MV3 extension should consume near-zero background memory because the worker is not running. On the downside, every wake-up re-parses the extension’s code, which shows up as CPU spikes and latency the first time you interact with a page after idling. If you only change one setting this year, make it this one: audit which of your extensions are still on Manifest V2 at chrome://extensions (Developer mode on, then read the version line). Chrome has been progressively disabling MV2 extensions since 2024, and any MV2 extension still running is on borrowed time.
Why “low memory” and “fast page load” are not the same measurement
A content script that injects 2 MB into every page can slow load times while barely registering in the Task Manager, because the cost lands in the page’s renderer process, not the extension’s. To measure this properly, use Chrome DevTools Performance panel with extensions enabled, record a page load, then repeat with all extensions disabled in an incognito window (extensions are off by default there). The delta in Total Blocking Time and DOMContentLoaded is your real extension tax.
Which extensions are worth their RAM in 2026
Ad blockers remain the only category with a credible claim to reducing total memory, because blocked ad scripts and trackers frequently outweigh the blocker’s own footprint. Everything else is a cost you pay for convenience.
- Keep unconditionally: one ad blocker (uBlock Origin or uBlock Origin Lite), one password manager (Bitwarden is the leaner default), Pocket, Notion Web Clipper.
- Keep but scope: Dark Reader (disable on sites that already have dark mode), Grammarly (restrict to text fields in settings), MetaMask (review connected sites quarterly).
- Enable on demand only: Loom, Zoom, Honey. These three are the most common cause of “why is Chrome using 4 GB” complaints, and all three work fine toggled on for a single task.
The spec sheet says a shopping extension is “lightweight,” but the measured behavior on retail pages — where these extensions inject price-comparison scripts — is where the cost appears. Test it yourself: open a product page with the extension on and off in incognito and compare Task Manager totals.
The hidden cost nobody mentions: extension count matters more than any single extension
Each installed extension adds a small fixed overhead — a background context, storage, and permission checks — even when idle. Twelve extensions at even 15 MB of idle overhead each is 180 MB before you open a single tab, which is roughly the memory footprint of a lightweight Linux desktop session in 2026.
Here is the practical rule: if an extension is not used weekly, uninstall it rather than disable it. Disabled extensions still occupy disk and appear in sync, and re-enabling is a 10-second task when you actually need one. The benchmark you should care about is not “which extension is heaviest” but “how many can I remove without changing my workflow.” For most readers, that answer is three to five.
How to run this benchmark on your own machine in ten minutes
Chrome’s Task Manager is the only authoritative source for your specific setup, and it takes three steps: press Shift+Esc, sort by the Memory column, and note the top five extension entries. Then open chrome://settings/performance and confirm Memory Saver is on — Google’s documentation states it frees memory from inactive tabs, though not from extension service workers.
For page-load impact, use DevTools → Performance → record a reload of a representative page (a news site and a retail product page are good choices), then repeat in incognito. Record Total Blocking Time for both runs. A difference above 200 ms is worth investigating; below that, you are within noise.
Verdict
Install one ad blocker, one password manager, and at most three other extensions you use weekly. Toggle Loom, Zoom, and any shopping extension on only when you need them, and uninstall rather than disable anything you have not touched in a month. The extension with the worst memory profile is not on any list — it is the one you forgot was installed. If you change a single thing after reading this, make it the MV2 audit at chrome://extensions; it is the highest-leverage five minutes you will spend on browser performance this year.